Predicting distant recurrence of endometrial cancer (EC) is crucial for personalized adjuvant treatment. The current gold standard of combined pathological and molecular profiling is costly,... Show morePredicting distant recurrence of endometrial cancer (EC) is crucial for personalized adjuvant treatment. The current gold standard of combined pathological and molecular profiling is costly, hampering implementation. Here we developed HECTOR (histopathology-based endometrial cancer tailored outcome risk), a multimodal deep learning prognostic model using hematoxylin and eosin-stained, whole-slide images and tumor stage as input, on 2,072 patients from eight EC cohorts including the PORTEC-1/-2/-3 randomized trials. HECTOR demonstrated C-indices in internal (n = 353) and two external (n = 160 and n = 151) test sets of 0.789, 0.828 and 0.815, respectively, outperforming the current gold standard, and identified patients with markedly different outcomes (10-year distant recurrence-free probabilities of 97.0%, 77.7% and 58.1% for HECTOR low-, intermediate- and high-risk groups, respectively, by Kaplan–Meier analysis). HECTOR also predicted adjuvant chemotherapy benefit better than current methods. Morphological and genomic feature extraction identified correlates of HECTOR risk groups, some with therapeutic potential. HECTOR improves on the current gold standard and may help delivery of personalized treatment in EC. Show less
Purpose: The primary objective of this study was to determine the feasibility of ablation margin quantification using a standardized scanning protocol during thermal ablation (TA) of hepatocellular... Show morePurpose: The primary objective of this study was to determine the feasibility of ablation margin quantification using a standardized scanning protocol during thermal ablation (TA) of hepatocellular carcinoma (HCC), and a rigid registration algorithm. Secondary objectives were to determine the inter- and intra-observer variability of tumor segmentation and quantification of the minimal ablation margin (MAM). Materials and methods: Twenty patients who underwent thermal ablation for HCC were included. There were thirteen men and seven women with a mean age of 67.1 +/- 10.8 (standard deviation [SD]) years (age range: 49.1-81.1 years). All patients underwent contrast-enhanced computed tomography examination under general anesthesia directly before and after TA, with preoxygenated breath hold. Contrast-enhanced computed tomography examinations were analyzed by radiologists using rigid registration software. Registration was deemed feasible when accurate rigid co-registration could be obtained. Inter- and intra-observer rates of tumor segmentation and MAM quantification were calculated. MAM values were correlated with local tumor progression (LTP) after one year of follow-up. Results: Co-registration of pre- and post-ablation images was feasible in 16 out of 20 patients (80%) and 26 out of 31 tumors (84%). Mean Dice similarity coefficient for inter- and intra-observer variability of tumor segmentation were 0.815 and 0.830, respectively. Mean MAM was 0.63 +/- 3.589 (SD) mm (range: -6.26 -6.65 mm). LTP occurred in four out of 20 patients (20%). The mean MAM value for patients who developed LTP was -4.00 mm, as compared to 0.727 mm for patients who did not develop LTP. Conclusion: Ablation margin quantification is feasible using a standardized contrast-enhanced computed tomography protocol. Interpretation of MAM was hampered by the occurrence of tissue shrinkage during TA. Further validation in a larger cohort should lead to meaningful cut-off values for technical success of TA. (c) 2023 The Author(s). Published by Elsevier Masson SAS on behalf of Societe francaise de radiologie. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/) Show less
Patients with acute coronary syndromes caused by plaque erosion might be managed conservatively without stenting. Currently, the diagnosis of plaque erosion requires an invasive imaging procedure.... Show morePatients with acute coronary syndromes caused by plaque erosion might be managed conservatively without stenting. Currently, the diagnosis of plaque erosion requires an invasive imaging procedure. We sought to develop a deep learning (DL) model that enables an accurate diagnosis of plaque erosion using coronary computed tomography angiography (CTA). A total of 532 CTA scans from 395 patients were used to develop a DL model: 426 CTA scans from 316 patients for training and internal validation, and 106 separate scans from 79 patients for validation. Momentum Distillation-enhanced Composite Transformer Attention (MD-CTA), a novel DL model that can effectively process the entire set of CTA scans to diagnose plaque erosion, was developed. The novel DL model, compared to the convolution neural network, showed significantly improved AUC (0.899 [0.841-0.957] vs. 0.724 [0.622-0.826]), sensitivity (87.1 [70.2-96.4] vs. 71.0 [52.0-85.8]), and specificity (85.3 [75.3-92.4] vs. 68.0 [56.2-78.3]), respectively, for the patient-level prediction. Similar results were obtained at the slice-level prediction AUC (0.897 [0.890-0.904] vs. 0.757 [0.744-0.770]), sensitivity (82.2 [79.8-84.3] vs. 68.9 [66.2-71.6]), and specificity (80.1 [79.1-81.0] vs. 67.3 [66.3-68.4]), respectively. This newly developed DL model enables an accurate CT diagnosis of plaque erosion, which might enable cardiologists to provide tailored therapy without invasive procedures.Clinical Trial Registration:http://www.clinicaltrials.gov, NCT04523194. Show less
BACKGROUND There are no data comparing sirolimus-coated balloons (SCBs [MagicTouch, Concept Medical]) to paclitaxel-coated balloons (PCBs [SeQuent Please Neo, B. Braun]) for the treatment of de... Show moreBACKGROUND There are no data comparing sirolimus-coated balloons (SCBs [MagicTouch, Concept Medical]) to paclitaxel-coated balloons (PCBs [SeQuent Please Neo, B. Braun]) for the treatment of de novo small vessel disease (SVD).OBJECTIVES This study sought to compare quantitative coronary angiographic outcomes at 6 months after treatment of de novo SVD with a PCB or SCB.METHODS This prospective, multicenter, noninferiority trial randomized 121 patients (129 SVD lesions) to treatment with an SCB or PCB, with balloon sizing determined using optical coherence tomography. The primary endpoint was noninferiority for the 6-month angiographic net lumen gain.RESULTS Angiographic follow-up was completed in 109 (90.1%) patients in the per-protocol analysis. The mean +/- SD angiographic net gains were 0.25 +/- 0.40 mm with SCBs vs 0.48 +/- 0.37 mm with PCBs, resulting in SCBs failing to meet the 0.30 mm criterion for noninferiority (P-noninferiority = 0.173), with an absolute difference of -0.23 mm (95% CI: -0.37 to -0.09) secondary to a smaller late loss (0.00 +/- 0.32 mm vs 0.32 +/- 0.47 mm; P < 0.001) and more frequent late lumen enlargement (53.7% vs 30.0%; OR: 2.60; 95% CI: 1.22-5.67; P = 0.014) with PCBs. Binary restenosis rates were 32.8% and 12.5% following treatment with SCBs and PCBs, respectively (OR: 3.41; 95% CI: 1.36-9.44; P = 0.012). The mean angiography-derived fractional flow ratio at follow-up was 0.86 +/- 0.15 following treatment with SCBs and 0.91 +/- 0.09 following PCBs (P = 0.026); a fractional flow ratio <= 0.80 occurred in 13 and 5 vessels after treatment with SCBs and PCBs, respectively.CONCLUSIONS The SCB MagicTouch failed to demonstrate noninferiority for angiographic net lumen gain at 6 months compared to the PCB SeQuent Please Neo. Show less
Driest, F.Y. van; Broersen, A.; Geest, R.J. van der; Jukema, J.W.; Scholte, A.J.H.A.; Dijkstra, J. 2023
Introduction: The use of serial coronary computed tomography angiography (CCTA) allows for the early assessment of coronary plaque progression, a crucial factor in averting major adverse cardiac... Show moreIntroduction: The use of serial coronary computed tomography angiography (CCTA) allows for the early assessment of coronary plaque progression, a crucial factor in averting major adverse cardiac events (MACEs). Traditionally, serial CCTA is assessed using anatomical landmarks to match baseline and follow-up scans. Recently, a tool has been developed that allows for the automatic quantification of local plaque thickness differences in serial CCTA utilizing plaque contour delineation.The aim of this study was to determine thresholds of plaque thickness differences that define whether there is plaque progression and/or regression. These thresholds depend on the contrast-to-noise ratio (CNR). Methods: Plaque thickness differences between two scans acquired at the same moment in time should always be zero. The negative and positive differences in plaque contour delineation in these scans were used along with the CNR in order to create calibration graphs on which a linear regression analysis was performed. This analysis was conducted on a cohort of 50 patients referred for a CCTA due to chest complaints. A total of 300 coronary vessels were analyzed. First, plaque contours were semi-automatically determined for all major epicardial coronary vessels. Second, manual drawings of seven regions of interest (ROIs) per scan were used to quantify the scan quality based on the CNR for each vessel. Results: A linear regression analysis was performed on the CNR and negative and positive plaque contour delineation differences. Accounting for the standard error of the estimate, the linear regression analysis revealed that above 1.009 - 0.002 9 CNR there is an increase in plaque thickness (progression), and below - 1.638 ? 0.012 9 CNR there is a decrease in plaque thickness (regression). Conclusion: This study demonstrates the feasibility of developing vessel-specific, qualitybased thresholds for visualizing local plaque thickness differences evaluated by serial CCTA. These thresholds have the potential to facilitate the early detection of atherosclerosis progression. Show less
Introduction: Changes in coronary artery luminal dimensions during the cardiac cycle can impact the accurate quantification of volumetric analyses in intravascular ultrasound (IVUS) image studies.... Show moreIntroduction: Changes in coronary artery luminal dimensions during the cardiac cycle can impact the accurate quantification of volumetric analyses in intravascular ultrasound (IVUS) image studies. Accurate ED-frame detection is pivotal for guiding interventional decisions, optimizing therapeutic interventions, and ensuring standardized volumetric analysis in research studies. Images acquired at different phases of the cardiac cycle may also lead to inaccurate quantification of atheroma volume due to the longitudinal motion of the catheter in relation to the vessel. As IVUS images are acquired throughout the cardiac cycle, end-diastolic frames are typically identified retrospectively by human analysts to minimize motion artefacts and enable more accurate and reproducible volumetric analysis.Methods: In this paper, a novel neural network-based approach for accurate end-diastolic frame detection in IVUS sequences is proposed, trained using electrocardiogram (ECG) signals acquired synchronously during IVUS acquisition. The framework integrates dedicated motion encoders and a bidirectional attention recurrent network (BARNet) with a temporal difference encoder to extract frame-by-frame motion features corresponding to the phases of the cardiac cycle. In addition, a spatiotemporal rotation encoder is included to capture the IVUS catheter's rotational movement with respect to the coronary artery.Results: With a prediction tolerance range of 66.7 ms, the proposed approach was able to find 71.9%, 67.8%, and 69.9% of end-diastolic frames in the left anterior descending, left circumflex and right coronary arteries, respectively, when tested against ECG estimations. When the result was compared with two expert analysts' estimation, the approach achieved a superior performance.Discussion: These findings indicate that the developed methodology is accurate and fully reproducible and therefore it should be preferred over experts for end-diastolic frame detection in IVUS sequences. Show less
Advances in image reconstruction using either single or multimodality imaging data provide increasingly accurate three-dimensional (3D) patient's arterial models for shear stress evaluation using... Show moreAdvances in image reconstruction using either single or multimodality imaging data provide increasingly accurate three-dimensional (3D) patient's arterial models for shear stress evaluation using computational fluid dynamics (CFD). We aim to evaluate the impacts on endothelial shear stress (ESS) derived from a simple image reconstruction using 3D-quantitative coronary angiography (3D-QCA) versus a multimodality reconstruction method using optical coherence tomography (OCT) in patients' vessels treated with bioresorbable scaffolds. Seven vessels at baseline and five-year follow-up of seven patients from a previous CFD investigation were retrospectively selected for a head-to-head comparison of angiography-derived versus OCT-derived ESS. 3D-QCA significantly underestimated the minimum stent area [MSA] (-2.38mm2) and the stent length (-1.46 mm) compared to OCT-fusion method reconstructions. After carefully co-registering the region of interest for all cases with a sophisticated statistical method, the difference in MSA measurements as well as the inability of angiography to visualise the strut footprint in the lumen surface have translated to higher angiography-derived ESS than OCT-derived ESS (1.76 Pa or 1.52 times for the overlapping segment). The difference in ESS widened with a more restricted region of interest (1.97 Pa or 1.63 times within the scaffold segment). Angiography and OCT offer two distinctive methods of ESS calculation. Angiography-derived ESS tends to overestimate the ESS compared to OCT-derived ESS. Further investigations into ESS analysis resolution play a vital role in adopting OCT-derived ESS. Show less
Intravascular ultrasound (IVUS) is recommended in guiding coronary intervention. The segmentation of coronary lumen and external elastic membrane (EEM) borders in IVUS images is a key step, but the... Show moreIntravascular ultrasound (IVUS) is recommended in guiding coronary intervention. The segmentation of coronary lumen and external elastic membrane (EEM) borders in IVUS images is a key step, but the manual process is time-consuming and error-prone, and suffers from inter-observer variability. In this paper, we propose a novel perceptual organisation-aware selective transformer framework that can achieve accurate and robust segmentation of the vessel walls in IVUS images. In this framework, temporal context-based feature encoders extract efficient motion features of vessels. Then, a perceptual organisation-aware selective transformer module is proposed to extract accurate boundary information, supervised by a dedicated boundary loss. The obtained EEM and lumen segmentation results will be fused in a temporal constraining and fusion module, to determine the most likely correct boundaries with robustness to morphology. Our proposed methods are extensively evaluated in non-selected IVUS sequences, including normal, bifurcated, and calcified vessels with shadow artifacts. The results show that the proposed methods outperform the state-of-the-art, with a Jaccard measure of 0.92 for lumen and 0.94 for EEM on the IVUS 2011 open challenge dataset. This work has been integrated into a software QCU-CMS1 to automatically segment IVUS images in a user-friendly environment. Show less
Safety and efficacy of coronary drug-eluting stents (DES) are often preclinically tested using healthy or minimally diseased swine. These generally show significant fibrotic neointima at follow-up,... Show moreSafety and efficacy of coronary drug-eluting stents (DES) are often preclinically tested using healthy or minimally diseased swine. These generally show significant fibrotic neointima at follow-up, while in patients, incomplete healing is often observed. The aim of this study was to investigate neointima responses to DES in swine with significant coronary atherosclerosis. Adult familial hypercholesterolemic swine (n = 6) received a high fat diet to develop atherosclerosis. Serial OCT was performed before, directly after, and 28 days after DES implantation (n = 14 stents). Lumen, stent and plaque area, uncovered struts, neointima thickness and neointima type were analyzed for each frame and averaged per stent. Histology was performed to show differences in coronary atherosclerosis. A range of plaque size and severity was found, from healthy segments to lipid-rich plaques. Accordingly, neointima responses ranged from uncovered struts, to minimal neointima, to fibrotic neointima. Lower plaque burden resulted in a fibrotic neointima at follow-up, reminiscent of minimally diseased swine coronary models. In contrast, higher plaque burden resulted in minimal neointima and more uncovered struts at follow-up, similarly to patients' responses. The presence of lipid-rich plaques resulted in more uncovered struts, which underscores the importance of advanced disease when performing safety and efficacy testing of DES. Show less
Multi-instance learning, a commonly used technique in artificial intelligence for analyzing slides, can be applied to diagnose thyroid cancer based on cytological smears. Since smears do not have... Show moreMulti-instance learning, a commonly used technique in artificial intelligence for analyzing slides, can be applied to diagnose thyroid cancer based on cytological smears. Since smears do not have mul-tidimensional histological features similar to histopathology, mining potential contextual information and diversity of features is crucial for better classification performance. In this paper, we propose a pyramid multi-loss vision transformer model called PyMLViT, a novel algorithm with two core modules to address these issues. Specifically, we design a pyramid token extraction module to acquire potential contextual information on smears. The pyramid token structure extracts multi-scale local features, and the vision transformer structure further obtains global information through the self -attention mechanism. Furthermore, we construct multi-loss fusion module based on the conventional multi-instance learning framework. With carefully designed bag and patch weight allocation strategies, we incorporate slide-level annotations as pseudo-labels for patches to participate in training, thus enhancing the diversity of supervised information. Extensive experimental results on the real-world dataset show that PyMLViT has a high performance and a competitive number of parameters compared to popular methods for diagnosing thyroid cancer in cytological smears.& COPY; 2023 Elsevier B.V. All rights reserved. Show less
Marvin, D.L.; Dijkstra, J.; Zulfiqar, R.M.; Vermeulen, M.; Dijke, P. ten; Ritsma, L. 2023
Despite advances in treatment for metastatic melanoma patients, patients with liver metastasis have an unfavorable prognosis. A better understanding of the development of liver metastasis is... Show moreDespite advances in treatment for metastatic melanoma patients, patients with liver metastasis have an unfavorable prognosis. A better understanding of the development of liver metastasis is needed. The multifunctional cytokine Transforming Growth Factor β (TGF-β) plays various roles in melanoma tumors and metastasis, affecting both tumor cells and cells from the surrounding tumor microenvironment. To study the role of TGF-β in melanoma liver metastasis, we created a model to activate or repress the TGF-β receptor pathway in vitro and in vivo in an inducible manner. For this, we engineered B16F10 melanoma cells to have inducible ectopic expression of a constitutively active (ca) or kinase-inactive (ki) TGF-β receptor I, also termed activin receptor-like kinase (ALK5). In vitro, stimulation with TGF-β signaling and ectopic caALK5 expression reduced B16F10 cell proliferation and migration. Contrasting results were found in vivo; sustained caALK5 expression in B16F10 cells in vivo increased the metastatic outgrowth in liver. Blocking microenvironmental TGF-β did not affect metastatic liver outgrowth of both control and caALK5 expressing B16F10 cells. Upon characterizing the tumor microenvironment of control and caALk5 expressing B16F10 tumors, we observed reduced (cytotoxic) T cell presence and infiltration, as well as an increase in bone marrow-derived macrophages in caALK5 expressing B16F10 tumors. This suggests that caALK5 expression in B16F10 cells induces changes in the tumor microenvironment. A comparison of newly synthesized secreted proteins upon caALK5 expression by B16F10 cells revealed increased secretion of matrix remodeling proteins. Our results show that TGF-β receptor activation in B16F10 melanoma cells can increase metastatic outgrowth in liver in vivo, possibly through remodeling of the tumor microenvironment leading to altered infiltration of immune cells. These results provide insights in the role of TGF-β signaling in B16F10 liver metastasis and could have implications regarding the use of TGF-β inhibitors for the treatment of melanoma patients with liver metastasis. Show less
Background: Advances in coronary computed tomography angiography (CCTA) reconstruction algorithms are expected to enhance the accuracy of CCTA plaque quantification. We aim to evaluate different... Show moreBackground: Advances in coronary computed tomography angiography (CCTA) reconstruction algorithms are expected to enhance the accuracy of CCTA plaque quantification. We aim to evaluate different CCTA recon-struction approaches in assessing vessel characteristics in coronary atheroma using intravascular ultrasound (IVUS) as the reference standard.Methods: Matched cross-sections (n = 7241) from 50 vessels in 15 participants with chronic coronary syndrome who prospectively underwent CCTA and 3-vessel near-infrared spectroscopy-IVUS were included. Twelve CCTA datasets per patient were reconstructed using two different kernels, two slice thicknesses (0.75 mm and 0.50 mm) and three different strengths of advanced model-based iterative reconstruction (IR) algorithms. Lumen and vessel wall borders were manually annotated in every IVUS and CCTA cross-section which were co-registered using dedicated software. Image quality was sub-optimal in the reconstructions with a sharper kernel, so these were excluded. Intraclass correlation coefficient (ICC) and repeatability coefficient (RC) were used to compare the estimations of the 6 CT reconstruction approaches with those derived by IVUS.Results: Segment-level analysis showed good agreement between CCTA and IVUS for assessing atheroma volume with approach 0.50/5 (slice thickness 0.50 mm and highest strength 5 ADMIRE IR) being the best (total atheroma volume ICC: 0.91, RC: 0.67, p < 0.001 and percentage atheroma volume ICC: 0.64, RC: 14.06, p < 0.001). At lesion-level, there was no difference between the CCTA reconstructions for detecting plaques (accuracy range: 0.64-0.67; p = 0.23); however, approach 0.50/5 was superior in assessing IVUS-derived lesion characteristics associated with plaque vulnerability (minimum lumen area ICC: 0.64, RC: 1.31, p < 0.001 and plaque burden ICC: 0.45, RC: 32.0, p < 0.001).Conclusion: CCTA reconstruction with thinner slice thickness, smooth kernel and highest strength advanced IR enabled more accurate quantification of the lumen and plaque at a segment-, and lesion-level analysis in coronary atheroma when validated against intravascular ultrasound. Clinicaltrials.gov (NCT03556644) Show less
Poon, E.K.W.; Ono, M.; Wu, X.L.; Dijkstra, J.; Sato, Y.; Kutyna, M.; ... ; Serruys, P.W. 2023
Endothelial shear stress (ESS) plays a key role in the clinical outcomes in native and stented segments; however, their implications in bypass grafts and especially in a synthetic biorestorative... Show moreEndothelial shear stress (ESS) plays a key role in the clinical outcomes in native and stented segments; however, their implications in bypass grafts and especially in a synthetic biorestorative coronary artery bypass graft are yet unclear. This report aims to examine the interplay between ESS and the morphological alterations of a biorestorative coronary bypass graft in an animal model. Computational fluid dynamics (CFD) simulation derived from the fusion of angiography and optical coherence tomography (OCT) imaging was used to reconstruct data on the luminal anatomy of a bioresorbable coronary bypass graft with an endoluminal "flap" identified during OCT acquisition. The "flap" compromised the smooth lumen surface and considerably disturbed the local flow, leading to abnormally low ESS and high oscillatory shear stress (OSI) in the vicinity of the "flap". In the presence of the catheter, the flow is more stable (median OSI 0.02384 versus 0.02635, p < 0.0001; maximum OSI 0.4612 versus 0.4837). Conversely, OSI increased as the catheter was withdrawn which can potentially cause back-and-forth motions of the "flap", triggering tissue fatigue failure. CFD analysis in this report provided sophisticated physiological information that complements the anatomic assessment from imaging enabling a complete understanding of biorestorative graft pathophysiology. Show less
Hajjar, L.A.; Ancona, M.B.; Kalil, R.; Tresoldi, M.; Caldas, J.G.; Monti, G.; ... ; Montorfano, M. 2023
Background Microvascular lung vessels obstructive thromboinflammatory syndrome has been proposed as a possible mechanism of respiratory failure in COVID-19 patients. However, it has only been... Show moreBackground Microvascular lung vessels obstructive thromboinflammatory syndrome has been proposed as a possible mechanism of respiratory failure in COVID-19 patients. However, it has only been observed in post-mortem studies and has never been documented in vivo, probably because of a lack of CT scan sensitivity in small pulmonary arteries. The aim of the present study was to assess the safety, tolerability, and diagnostic value of optical coherence tomography (OCT) for the assessment of patients with COVID-19 pneumonia for pulmonary microvascular thromboinflammatory syndrome.Methods The COVID-OCT trial was a multicenter, open-label, prospective, interventional clinical study. Two cohorts of patients were included in the study and underwent pulmonary OCT evaluation. Cohort A consisted of patients with COVID-19 with a negative CT scan for pulmonary thrombosis and elevated thromboinflammatory markers (D-dimer > 10,000 ng/mL or 5,000 < D-dimer < 10,000 ng/mL and one of: C-reactive Protein > 100 mg/dL, IL-6 > 6 pg/mL, or ferritin > 900 ng/L). Cohort B consisted of patients with COVID-19 and a CT scan positive for pulmonary thrombosis. The primary endpoints of the study were: (i) to evaluate the overall safety of OCT investigation in patients with COVID-19 pneumonia, and (ii) to report on the potential value of OCT as a novel diagnostic tool for the diagnosis of microvascular pulmonary thrombosis in COVID-19 patients.Results A total of 13 patients were enrolled. The mean number of OCT runs performed in each patient was 6.1 +/- 2.0, both in ground glass and healthy lung areas, achieving a good evaluation of the distal pulmonary arteries. Overall, OCT runs identified microvascular thrombosis in 8 patients (61.5%): 5 cases of red thrombus, 1 case of white thrombus, and 2 cases of mixed thrombus. In Cohort A, the minimal lumen area was 3.5 +/- 4.6 mm(2), with stenosis of 60.9 +/- 35.9% of the area, and the mean length of thrombus-containing lesions was 5.4 +/- 3.0 mm. In Cohort B, the percentage area obstruction was 92.6 & PLUSMN; 2.6, and the mean thrombus-containing lesion length was 14.1 +/- 13.9 mm. No peri-procedural complications occurred in any of the 13 patients.Conclusion OCT appears to be a safe and accurate method of evaluating the distal pulmonary arteries in hospitalized COVID-19 patients. Here, it enabled the first in vivo documentation of distal pulmonary arterial thrombosis in patients with elevated thromboinflammatory markers, even when their CT angiogram was negative for pulmonary thrombosis. Show less
AimsEarly healing after drug-eluting stent (DES) implantation may reduce the risk of stent thrombosis. The aim of this study was to compare patterns of early healing after implantation of the thin... Show moreAimsEarly healing after drug-eluting stent (DES) implantation may reduce the risk of stent thrombosis. The aim of this study was to compare patterns of early healing after implantation of the thin strut everolimus-eluting Synergy DES (Boston Scientific) or the biolimus-eluting Biomatix Neoflex DES (Biosensors). Methods and ResultsA total of 160 patients with the chronic or acute coronary syndrome were randomized 1:1 to Synergy or Biomatrix DES. Optical coherence tomography (OCT) was performed at baseline and at either 1- or 3-month follow-up. The primary endpoint was a coronary stent healing index (CSHI), a weighted index of strut coverage, neointimal hyperplasia, malapposition, and extrastent lumen. A total of 133 cases had OCT follow-up and 119 qualified for matched OCT analysis. The median CSHI score did neither differ significantly between the groups at 1 month: Synergy 8.0 (interquartile range [IQR]: 3.0; 14.0) versus Biomatrix 8.5 (IQR: 4.0; 15.0) (p = 0.47) nor at 3 months: Synergy 6.5 (IQR: 2.0; 13.0) versus Biomatrix 6.0 (IQR: 4.0; 11.0) (p = 0.83). Strut coverage was 84.6% (IQR: 72.0; 97.9) for Synergy versus 77.6% (IQR: 70.1; 90.3) for Biomatrix (p = 0.15) at 1 month and 90.3% (IQR 79.0; 98.8) (Synergy) versus 83.9% (IQR: 77.5; 92.6) (Biomatrix) (p = 0.068) at 3 months. Pooled 1- and 3-month coverage was 88.6% (IQR: 74.4; 98.4) for Synergy compared with 80.7% (IQR: 73.2; 90.8) for Biomatrix (p = 0.02). ConclusionsThe early healing response after treatment with the Synergy or Biomatrix DES did not differ significantly as determined by a healing index. The Synergy DES showed overall better early stent strut coverage. Show less
Background Endometrial cancer can be molecularly classified into POLEmut , mismatch repair deficient (MMRd), p53 abnormal (p53abn), and no specific molecular profile (NSMP) subgroups. We aimed to... Show moreBackground Endometrial cancer can be molecularly classified into POLEmut , mismatch repair deficient (MMRd), p53 abnormal (p53abn), and no specific molecular profile (NSMP) subgroups. We aimed to develop an interpretable deep learning pipeline for whole-slide-image-based prediction of the four molecular classes in endometrial cancer (im4MEC), to identify morpho-molecular correlates, and to refine prognostication. Methods This combined analysis included diagnostic haematoxylin and eosin-stained slides and molecular and clinicopathological data from 2028 patients with intermediate-to-high-risk endometrial cancer from the PORTEC-1 (n=466), PORTEC-2 (n=375), and PORTEC-3 (n=393) randomised trials and the TransPORTEC pilot study (n=110), the Medisch Spectrum Twente cohort (n=242), a case series of patients with POLEmut endometrial cancer in the Leiden Endometrial Cancer Repository (n=47), and The Cancer Genome Atlas-Uterine Corpus Endometrial Carcinoma cohort (n=395). PORTEC-3 was held out as an independent test set and a four-fold cross validation was performed. Performance was measured with the macro and class-wise area under the receiver operating characteristic curve (AUROC). Whole-slide images were segmented into tiles of 360 & mu;m resized to 224 x 224 pixels. im4MEC was trained to learn tile-level morphological features with self-supervised learning and to molecularly classify whole-slide images with an attention mechanism. The top 20 tiles with the highest attention scores were reviewed to identify morpho-molecular correlates. Predictions of a nuclear classification deep learning model serve to derive interpretable morphological features. We analysed 5-year recurrence-free survival and explored prognostic refinement by molecular class using the Kaplan-Meier method. Findings im4MEC attained macro-average AUROCs of 0 & BULL;874 (95% CI 0 & BULL;856-0 & BULL;893) on four-fold cross-validation and 0 & BULL;876 on the independent test set. The class-wise AUROCs were 0 & BULL;849 for POLEmut (n=51), 0 & BULL;844 for MMRd (n=134), 0 & BULL;883 for NSMP (n=120), and 0 & BULL;928 for p53abn (n=88). POLEmut and MMRd tiles had a high density of lymphocytes, p53abn tiles had strong nuclear atypia, and the morphology of POLEmut and MMRd endometrial cancer overlapped. im4MEC highlighted a low tumour-to-stroma ratio as a potentially novel characteristic feature of the NSMP class. 5-year recurrence-free survival was significantly different between im4MEC predicted molecular classes in PORTEC-3 (log-rank p<0 & BULL;0001). The ten patients with aggressive p53abn endometrial cancer that was predicted as MMRd showed inflammatory morphology and appeared to have a better prognosis than patients with correctly predicted p53abn endometrial cancer (p=0 & BULL;30). The four patients with NSMP endometrial cancer that was predicted as p53abn showed higher nuclear atypia and appeared to have a worse prognosis than patients with correctly predicted NSMP (p=0 & BULL;13). Patients with MMRd endometrial cancer predicted as POLEmut had an excellent prognosis, as do those with true POLEmut endometrial cancer. Interpretation We present the first interpretable deep learning model, im4MEC, for haematoxylin and eosin-based prediction of molecular endometrial cancer classification. im4MEC robustly identified morpho-molecular correlates and could enable further prognostic refinement of patients with endometrial cancer. Copyright & COPY; 2022 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license. Show less
Huang, J.Y.; Tu, S.X.; Masuda, S.; Ninomiya, K.; Dijkstra, J.; Chu, M.; ... ; Wijns, W. 2022
ObjectivesThe objective of the present study was to compare plaque burden (PB) calculated from optical coherence tomography (OCT) using deep learning (DL) with PB derived from co-registered... Show moreObjectivesThe objective of the present study was to compare plaque burden (PB) calculated from optical coherence tomography (OCT) using deep learning (DL) with PB derived from co-registered intravascular ultrasound (IVUS). BackgroundA DL algorithm was developed for automated plaque characterization and PB quantification from OCT images. However, the performance of this algorithm for PB quantification has not been validated. MethodsFive-year follow-up OCT and IVUS images from 15 patients implanted with bioresorbable vascular scaffold (BVS) at baseline were analyzed. Precise co-registration for 72 anatomical slices was achieved utilizing unique BVS radiopaque markers. PB derived from OCT DL and IVUS were compared. OCT cross-sections were divided into four subgroups with different media visibility level. The impact of media visibility on the numerical difference between OCT-derived and IVUS-derived PB was investigated. The stent sizes selected by OCT DL and IVUS were compared. ResultsSixty-four paired OCT and IVUS cross-sections were compared. OCT DL showed good concordance with IVUS for PB assessment (ICC = 0.81, difference = -3.53 +/- 6.17%, p < 0.001). The numerical difference between OCT DL-derived PB and IVUS-derived PB was not substantially impacted by missing segments of media visualization (p = 0.21). OCT DL showed a diagnostic accuracy of 92% in identifying PB > 65%. The stent sizes selected by OCT DL were smaller compared to the ones selected by IVUS (difference = 0.30 +/- 0.34 mm, p < 0.001). ConclusionsThe DL algorithm provides a feasible and reliable method for automated PB estimation from OCT, irrespective of media visibility. OCT DL showed good diagnostic accuracy in identifying PB > 65%, revealing its potential to complement conventional OCT imaging. Show less
Purpose: To assess the quantitative minimal ablation margin (MAM) in patients with colorectal liver metastases (CRLM) treated with percutaneous thermal ablation (TA) and correlate the quantitative... Show morePurpose: To assess the quantitative minimal ablation margin (MAM) in patients with colorectal liver metastases (CRLM) treated with percutaneous thermal ablation (TA) and correlate the quantitative MAM with local tumour recurrence (LTR).Method: Thirty-nine of 143 patients with solitary or multiple CRLM who underwent a first percutaneous TA procedure between January 2011 and May 2020 were considered eligible for study enrolment. Image fusion of pre-and post-ablation scans and 3D quantitative MAM assessment was performed using the in-house developed semi-automatic rigid MRI/CECT-CECT co-registration software deLIVERed. The quantitative MAM was analysed and correlated with LTR.Results: Eighteen (46 %) patients were additionally excluded from further analyses due to suboptimal co -registration (quality co-registration score < 3). The quality of co-registration was considered sufficient in 21 (54 %) patients with a total of 29 CRLM. LTR was found in 5 of 29 (17 %) TA-treated CRLM. In total, 12 (41 %) negative MAMs were measured (mean MAM-4.7 +/- 2.7 mm). Negative MAMs were significantly more frequently seen in patients who developed LTR (100 %) compared to those without LTR (29 %; p = 0.003). The median MAM of patients who developed LTR (-6.6 mm (IQR-9.5 to-4.6)) was significantly smaller compared to the median MAM of patients without LTR (0.5 mm (IQR-1.8 to 3.0); p < 0.001). The ROC curve showed high accuracy in predicting LTR for the quantitative MAM (area under the curve of 0.975 +/- 0.029).Conclusion: This study demonstrated the feasibility of 3D quantitative MAM assessment, using deLIVERed co -registration software, to assess technical success of TA in patients with CRLM and to predict LTR. Show less
Vascular inflammation, lipid metabolism, and thrombogenicity play a key role not only in atherogenesis but also in the development of acute coronary syndromes. Biomarkers associated with coronary... Show moreVascular inflammation, lipid metabolism, and thrombogenicity play a key role not only in atherogenesis but also in the development of acute coronary syndromes. Biomarkers associated with coronary high-risk plaques defined according to intravascular imaging have not been systematically studied. A total of 69 patients with coronary artery disease who underwent both optical coherence tomography and intravascular ultrasound imaging, and who provided blood specimens were included. Comprehensive biomarkers for inflammation, lipid, and coagulation were analyzed. Composite models sought biomarker patterns associated with thin-cap fibroatheroma (TCFA) and "high-risk plaques" (TCFA and large plaque burden). Two different composite models were developed for TCFA, based on the finding that high sensitivity C-reactive protein (hsCRP), plasminogen activator inhibitor-1, fibrinogen, IL-6, homocysteine and amyloid A levels were elevated, and high-density lipoprotein cholesterol (HDL) and bile acid levels were decreased in these patients. Both composite models were highly accurate for detecting patients with TCFA (area under curve [AUC]: 0.883 in model-A and 0.875 in model-B, both p < 0.001). In addition, creatinine, hsCRP, fibrinogen, tumor necrosis factor-a, IL-6, homocysteine, amyloid A, HDL, prothrombin, and bile acid were useful for detecting patients with "high-risk plaques". Two composite models were highly accurate for detection of patients with "high-risk plaques" (AUC: 0.925 in model-A and 0.947 in model-B, both p <0.001). Biomarkers useful for detection of patients with high-risk coronary plaques defined according to intravascular imaging have been identified. These biomarkers may be useful to risk stratify patients and to develop targeted therapy. Show less
BACKGROUND: The relationship between gut microbiota and in vivo coronary plaque characteristics has not been reported. This study was conducted to investigate the relationship between gut... Show moreBACKGROUND: The relationship between gut microbiota and in vivo coronary plaque characteristics has not been reported. This study was conducted to investigate the relationship between gut microbiota and coronary plaque characteristics in patients with coronary artery disease.METHODS AND RESULTS: Patients who underwent both optical coherence tomography and intravascular ultrasound imaging and provided stool and blood specimens were included. The composition of gut microbiota was evaluated using 16S rRNA sequencing. A total of 55 patients were included. At the genus level, 2 bacteria were associated with the presence of thin-cap fibroatheroma, and 9 bacteria were associated with smaller fibrous cap thickness. Among them, some bacteria had significant associations with inflammatory/prothrombotic biomarkers. Dysgonomonas had a positive correlation with interleukin-6, Paraprevotella had a positive correlation with fibrinogen and negative correlation with high-density lipoprotein cholesterol, Succinatimonas had positive correlations with fibrinogen and homocysteine, and Bacillus had positive correlations with fibrinogen and high-sensitivity C-reactive protein. In addition, Paraprevotella, Succinatimonas, and Bacillus were also associated with greater plaque volume. Ten bacteria were associated with larger fibrous cap thickness. Some were associated with protective biomarker changes; Anaerostipes had negative correlations with trimethylamine N-oxide, tumor necrosis factor alpha, and interleukin-6, and Dielma had negative correlations with trimethylamine N-oxide, white blood cells, plasminogen activator inhibitor-1, and homocysteine, and a positive correlation with high-density lipoprotein cholesterol.CONCLUSIONS: Bacteria that were associated with vulnerable coronary plaque phenotype and greater plaque burden were identified. These bacteria were also associated with elevated inflammatory or prothrombotic biomarkers. Show less
